Transgenic Myocardial Overexpression of Fibroblast Growth Factor-1 Increases Coronary Artery Density and Branching
Abstract—Fibroblast growth factor (FGF)-1 plays important roles during myocardial and coronary morphogenesis. FGF-1 is also involved in the physiological response of the adult heart against ischemia, which includes cardiomyocyte protection and vascular growth. In the present study, we have generated transgenic mice with specific myocardial overexpression of the gene. Transgene expression was verified by Northern blot, and increased FGF-1 protein content was assessed by Western blot and immunoconfocal microscopy. Anatomic, histomorphological, and ultrastructural analyses revealed no major morphological or developmental abnormalities of transgenic hearts. Capillary density was unaltered, whereas the density of coronary arteries, especially arterioles, was significantly increased, as was the number of branches of the main coronary arteries. In addition, the coronary flow was significantly enhanced in transgenic mice ex vivo. These differences in the anatomic pattern of the coronary vasculature are established during the second month of postnatal life. The present findings demonstrate an important role of FGF-1 in the differentiation and growth of the coronary system and suggest that it is a key regulatory molecule of the differentiation of the arterial system.
Fibroblast growth factor (FGF)-1 and -2 are potent angiogenic factors inducing capillary formation.1 FGF-1 and -2 also induce vascular smooth muscle cell (SMC) proliferation and migration in cell cultures.2 3 In vivo, FGF-1 mRNA is normally expressed by cardiomyocytes, and the protein is located in myocytes and in the extracellular matrix.4 5
During embryonic development, FGF is crucial for the differentiation of mesodermal derivatives, including the heart.6 7 In addition, the expression and localization of FGF-1 ligand and receptors during cardiac morphogenesis correlate with the proliferation and differentiation of endothelial and SMCs of the coronary vessels, as well as of cardiomyocytes.5 8 9
In the adult heart, the FGF receptor and ligand have been demonstrated to be upregulated during ischemia in several animal models.10 11 In addition, local or systemic administration of FGF during episodes of myocardial ischemia improves collateral flow and left ventricular function and induces new vessel formation and SMC hyperplasia in arterioles and small arteries.12 13 14 15 16
All these studies indicate that FGF-1 is implicated in the proliferation and differentiation of most of the cellular components of the embryonic heart, as well as in the vascular growth associated with ischemia. To study these effects in detail, we generated transgenic (TR) mice with specific myocardial overexpression of the human FGF-1 cDNA.
Materials and Methods
Animals were handled in accordance with the American Physiological Society guidelines for animal welfare and the Bioethical Committee of the District of Darmstadt, Germany. Animals were housed in standard cages and fed water and chow ad libitum.
Generation of FGF-1–Overexpressing Mice
A 2.2-kb EcoRI fragment of clone pHBGF1.3A, including the coding and 3′ untranslated sequence of human FGF-1 but lacking any signal sequence (kindly provided by Dr I.M. Chiu, Ohio State University, Columbus), was ligated between the SV40 large intron and the 2.2-kb mouse MLC2v promoter (plasmid MLC2v/FGF), with the latter providing cardiac-specific expression (line 7 [L7]). In another construct, the CMV enhancer was added upstream from the promoter17 (plasmid CMV/MLC2v/FGF, line 1 [L1]) (see Figure 1A⇓).
The constructs were microinjected into the male pronucleus of murine CD2F1 zygotes and transferred into pseudopregnant females. The presence of transgenes was tested either by Southern blot or by genomic polymerase chain reaction. Three TR lines for each construct were generated.
Northern blot was performed on the heart, liver, kidney, lung, and skeletal muscles of 12- to 16-week-old mice, as previously described18 (see also online Materials and Methods, available at http://www.circresaha.org).
Western blot was performed on hearts from TR mice and their nontransgenic littermates (WT mice) by use of polyclonal goat anti–FGF-1 antibodies (Santa Cruz) (for details see online Materials and Methods, available at http://www.circresaha.org). Recombinant human FGF-1 (Santa Cruz) served as a positive control, and omission of the first antibody served as a negative control.
Gross Anatomy and Histomorphological Analysis
The hearts of WT and TR mice were dissected, and the gross anatomy was assessed under the binocular microscope. Alternatively, the hearts were cryopreserved or perfusion-fixed and embedded in paraffin.
The following antibodies were used: polyclonal antibody against FGF-1 (Promega), polyclonal antibody against the proliferation marker Ki-67 (Dianova), and FITC-conjugated monoclonal antibodies against vascular smooth muscle α-actin (Sigma Chemical Co). The TRITC-conjugated lectin Bandeiraea simplicifolia (BS-I, Sigma) was also used. Immunofluorescence was performed in 10-μm paraffin and cryostat sections as previously described19 (see also online Materials and Methods, available at http://www.circresaha.org).
The measurements of immunofluorescence intensity were performed with a Leica TCSNT confocal microscope as described elsewhere.20 Three-dimensional immunofluorescence was performed in 100-μm-thick vibratome sections as previously described.21
Hearts were perfusion-fixed, immersed in glutaraldehyde, and postfixed with osmium tetroxide. After they were rinsed in a series of ethanol, the samples were embedded in epoxy resin by following routine methods. Thin sections were poststained with uranyl acetate and Reynolds lead citrate and photographed with a Philips CM10 electron microscope.
Hearts were quickly dissected, cannulated through the ascending aorta, and retrogradely perfused with Krebs’ buffer supplemented with 0.01% adenosine. The coronary perfusate was collected, and the volume was measured. Finally, the hearts were weighed. Preliminary experiments showed that addition of BSA or any other protein to the perfusion buffer was not necessary.
The Bonferroni t test, paired t test, or ANOVA was used to examine the differences between experimental groups. All data were presented as mean±SD. Quantitative analyses were performed with the evaluators blinded to genotype.
An expanded Materials and Methods section is available online at http://www.circresaha.org.
Generation of TR Animals and Molecular Biology
Two lines, one for each construct, ie, L1 (CMV/MLC2v/FGF) and L7 (MLC2v/FGF), were analyzed in the present study (Figure 1A⇑). Both lines showed overexpression of the transgene on the RNA level, which was indicated by the appearance of a transgene-specific 2.1-kb mRNA in the heart (Figures 1B⇑ and 1C⇑). In WT animals, only the endogenous mRNA was expressed at very weak levels. No expression of the transgene mRNA was observed in other organs (not shown).
Western blot analysis confirmed overexpression (1.8-fold increase, data not shown) of the transgene at the protein level in the hearts of TR animals compared with WT animals (Figure 1D⇑). Note that the antibodies used could not differentiate between human and mouse FGF-1.
Immunoconfocal microscopy revealed localization of FGF-1 in the cardiomyocytes and extracellular matrix of both TR and WT hearts (Figures 2A⇓ through 2D). The pattern of immunostaining was similar in both groups of animals. However, TR mice showed a significant increase in the FGF-1 expression levels, especially in subepicardial areas of the myocardium, compared with WT mice (subepicardium, 1.99-fold; subendocardium, 1.34-fold) (Figure 2E⇓).
Gross Anatomy and Histomorphology
The size of the cardiac chambers was similar in TR and WT mice (Figures 3A⇓ and 3B⇓). In TR specimens, ventricular trabeculation, papillary muscles, cardiac valves, and atrioventricular and interventricular septa appeared anatomically and histologically intact.
The origin, distribution, and localization of the coronary arteries were similar in TR and WT specimens as assessed by serial paraffin sections of the whole heart (data not shown).
Electron microscopy revealed that cardiomyocytes in both WT and TR mice showed normal nuclei, intact sarcolemma, numerous mitochondria with densely packed cristae, and myofilaments with typical Z and A bands (Figures 4A⇓ and 4B⇓). In TR mice, small arterioles with one layer of SMCs showed a morphology not different from that of control mice (Figures 4C⇓ and 4D⇓). The main coronary arteries in either WT or TR hearts were composed of 5 or 6 layers of SMCs with the contractile phenotype (Figures 4E⇓ and 4F⇓). Endothelial cells were of flat shape and separated from SMCs by a continuous elastic lamina.
The thickness of the interventricular septum was measured in paraffin sections stained with hematoxylin and eosin. No significant differences between TR and WT animals were detected (Figure⇑ 3C).
The density of coronary capillary vessels was calculated by using the endothelium-specific marker BS-1. Only transversally cut capillary vessels were counted. No significant differences between TR and WT specimens were detected (Figure 5A⇓). The quantification was repeated by different observers with similar results (data not shown).
The density of immunodetectable α-actin–positive coronary arteries was calculated by using cryostat and paraffin sections in separate experimental settings. Veins were recognized by their morphology and excluded. The density of coronary arteries was increased in TR mice (L1, 1.45-fold; L7, 1.29-fold) compared with WT mice (Figure 5B⇑).
The internal diameter of the coronary arteries was measured in paraffin sections at similar distal and proximal levels of the heart. Only arteries with a circular or oval perimeter were measured. The density of arteries belonging to 4 different size groups was compared between TR and WT specimens (Figure 5C⇑). The greatest difference in numerical density (1.36-fold) was obtained in arteries with <30 μm of diameter (arterioles). The main coronary arteries (>90 μm) and branches of the first order (60 to 90 μm) also showed significant differences (Figure 5C⇑) that were probably due to the small number of vessels present in each section. However, these 2 groups represent a very small proportion of the total arterial density (60 to 90 μm, ≈1%; >90 μm, ≈10.5%).
The total number of branches of the left main coronary artery was counted in serial paraffin sections of the heart and was made relative to the total length of the artery, from its proximal origin in the aorta to the last branch in the apex of the heart. The relative number of branches was significantly increased in TR animals compared with WT animals (L1, 1.5-fold; L7, 1.4-fold) (Figure 5D⇑). Despite this difference, the analysis of serial histological sections revealed no alteration in the distribution of the arterial branches.
Figures 6A⇓ and 6B⇓ are 3D immunoconfocal reconstructions of single arteriolar trees labeled for α-actin. Small arteries of similar size show a higher number of arteriolar branches in TR animals, leading to a more complex arteriolar vascular bed than is found in WT animals.
The proliferation rate was calculated in cryosections of adult TR and WT hearts. TR animals showed a 1.67-fold significant increase in the total number of proliferating cells (Figure 7A⇓). Most of the proliferating cells were identified as interstitial cells, although no evident signs of fibrosis were found in TR hearts. Double labeling using Ki-67 and α-actin antibodies revealed a negligible number of proliferating cells in coronary arteries or arterioles (data not shown).
To investigate the increase in arterial density detected in TR hearts by morphometric techniques, we performed ex vivo hemodynamic experiments, with the hypothesis that a higher density of arterioles would increase the coronary flow under maximal vasodilatory conditions. The coronary flow was significantly higher in TR specimens than in WT specimens for every pressure tested (Figure 7B⇑). The flow differences gradually increased with the pressure used, from 1.19-fold at 80 mm Hg to 1.25-fold at 142 mm Hg. The slope of the regression function corresponding to the TR group was higher (1.36-fold) than that of the WT group (Figure 7C⇑). The slope of the regression curve for change in flow (0.004) was significantly different from 0 (Figure⇑ 7C), indicating that the pressure-dependent increase in the coronary flow was elevated in TR animals.
Morphometry of Postnatal Hearts
The density of coronary arteries was calculated in postnatal TR (L1) and WT mice as described above. The numerical density of coronary arteries decreased in both TR and WT groups from 2 weeks of postnatal age until adulthood (36 weeks) (Figure 8⇓), a fact that can be attributed to the fast postnatal growth of the ventricular mass compared with the growth of the vasculature.22 However, the decrease in arterial density was less accentuated in TR animals, leading to a maximal difference (1.4-fold) in adult TR hearts compared with WT hearts. The coronary artery density was also significantly increased in 6-week-old (1.25-fold) and 10-week-old (1.12-fold) TR animals.
We have generated TR mice with cardiac-specific overexpression of the human FGF-1 gene. TR mice were viable and healthy and showed a normal life span, indicating that the transgene caused no major congenital malformation.
Transgene mRNA expression was demonstrated in the hearts of TR mice but not of WT mice (Figure 1B⇑). TR animals belonging to L1 showed a higher level of transgene mRNA than did L7 animals, indicating an improvement of the expression mediated by the CMV enhancer. We also noted a 6-fold increase of the endogenous FGF-1 mRNA in TR animals of L1, the reason of which is unknown but may be triggered by the CMV enhancer of the transgene.
At the protein level, Western blot analyses and immunofluorescence measurements showed a significant increase of FGF-1 in TR hearts compared with WT hearts (Figures 1C⇑ and 2⇑). The protein was localized in cardiomyocytes, the extracellular matrix, and some interstitial cells, with an epicardial to endocardial gradient of expression. The increase in the immunofluorescence intensity of FGF-1 in TR animals was particularly evident in the subepicardial areas of the myocardium, a fact that can be interpreted as a differential anatomic regulation of MLC2v promoter expression. However, to our knowledge, such a differential expression has not been reported to date.
During embryonic development, FGF-1 and FGF receptor-1 are abundantly expressed in cardiomyocytes, correlating with the proliferation and differentiation stages of cardiogenesis.5 6 8 Moreover, in vivo retroviral transfection of embryonic chicken myocardium with a dominant-negative mutant of FGF receptor-1 reduces the proliferation activity of cardiomyocytes,6 demonstrating the mitogenic capacity of FGF. However, our anatomic, histological, and ultrastructural analyses of FGF-1 TR mice revealed no morphological alterations of the myocardium, indicating that FGF-1 overexpression had no effect on cardiomyocyte development and growth. This fact may be explained as a receptor-mediated regulation of the FGF-1 function in cardiomyocytes. Alternatively, members of the FGF family other than FGF-1 might trigger differentiation and proliferation of cardiomyocytes during embryonic development.
The quantification of the capillary density in the 2 TR lines analyzed showed values similar to those of WT mice, indicating that angiogenesis is not perturbed in this TR model (Figure 5A⇑). However, we found a significant 1.45-fold increase in the density of arteries in L1 TR mice compared with WT mice and a 1.29-fold increase in L7 TR mice (Figure 5B⇑). In addition, the number of branches of the coronary arteries was significantly increased in both TR lines compared with WT mice (1.45-fold and 1.40-fold, respectively) (Figure 5D⇑). This altered vascular pattern can be attributed to an increase in the number of small arterioles, as indicated by the quantification of the density of arteries relative to the vessel size (Figures 5C⇑ and 6⇑). The ex vivo hemodynamic study confirmed the morphological results, showing an enhanced coronary flow in the hearts of TR mice (Figures 7B⇑ and 7C⇑). This enhanced flow was not due to the vasoactive capacity of FGF because all the experiments were performed under maximal vasodilatory conditions. Moreover, the differences in coronary flow between TR and WT specimens increased with the perfusion pressure, resulting in a significant elevation of the pressure-dependent increase in the coronary flow of TR animals. In summary, morphological and functional data demonstrate a moderate overgrowth of the resistance vessels in the hearts of FGF-1 TR mice, indicating a role for FGF-1 in the differentiation and/or growth of the coronary arterial system but not in the capillary network.
Several studies on therapeutic angiogenesis support the hypothesis that FGF-1 plays a role in the growth of arteries. Administration of native or recombinant FGF-1 or of vectors coding for FGF-1 in ischemic and in nonischemic tissues causes growth of capillaries and arteries.8 23 24 25 These studies agree with our observations of a strong influence of FGF-1 on the growth of the arterial system. The question now is which is the biological mechanism, mediated by FGF-1, responsible for the formation, differentiation, and/or growth of new arterioles in TR hearts.
Despite the increased arteriolar density found in our TR mice, the proliferation rate of endothelial and SMCs was not altered in adult TR mice compared with WT mice. One possible explanation is that the arteriole overgrowth does not take place in the adult and fully differentiated heart but during fetal or postnatal life. Indeed, our morphometric studies indicate that the numerical difference in arterial density between TR and WT mice appears first in animals between 2 and 6 weeks of postnatal age (Figure⇑ 8), indicating that the FGF-1–dependent coronary artery overgrowth of TR animals occurs during the first postnatal month. Moreover, studies in the rat heart indicate that the arteriolar growth stops by postnatal week 426 ; this growth arrest is correlated with the downregulation of FGF-1 in arterial SMCs.5 We propose that abnormal levels of FGF-1 in the heart of postnatal TR mice may maintain the growth status of the arterial system, leading to the coronary artery overgrowth observed in FGF-1 TR animals.
The fact that the relative number of branches of the coronary arteries is increased in FGF-1 TR mice has special relevance. Dilley and Schwartz27 described in 1989 the altered structure of the arterial tree in the growth hormone TR mouse. These extremely big mice must adapt their vascular system to carry twice the normal blood flow. The adaptation includes wider lumens and thicker vessel walls, but the arterial tree branches the same number of times as in the normal mice. These studies indicate that the vasculature contains a genetically determined circuit plan, which is not dependent on environmental cues, a statement that has been already proposed by other authors.28 The finding of a 1.45-fold increase in the number of arterial branches in FGF-1 TR mice indicates an alteration of the genetically determined circuit plan in this mouse model and strongly supports the hypothesis that this growth factor is a basic regulator of the structure of arterial trees. Interestingly, FGF-7 has been demonstrated to induce branching morphogenesis of the prostate epithelium during the embryonic development.29 Moreover, one isoform of the FGF receptor-2 (FGFR-2 IIIb) is normally expressed in the branch points of the bronchioles of the developing mammalian lung, and the expression of a dominant-negative form of this receptor results in a complete loss of branching morphogenesis.30 Both FGF-1 and FGF-7 can bind to FGFR-2 IIIb.31 It is tempting to speculate that different members of the FGF gene family exert similar biological functions, with regard to the regulation of the branching morphogenesis of different embryonic structures.
In summary, our results demonstrate a role for FGF-1 in the differentiation and/or growth of the coronary arterial system during postnatal life. The unique phenotype of FGF-1 TR mice, with a quantitative alteration of the number of the resistance vessels and their branches within the background of an anatomically, histologically, and ultrastructurally normal vasculature, suggests that FGF-1 acts as a master regulatory gene in the differentiation of the arterial system.
The authors thank Prof Dr Jutta Schaper for her help with morphological techniques and Beate Grohmann and Claudia Ullmann for technical assistance.
- Received May 11, 2000.
- Accepted June 8, 2000.
- © 2000 American Heart Association, Inc.
Gospodaroviwicz D, Neufeld G, Schweigerer L. Molecular and biological characterization of FGF, an angiogenic factor which also controls the proliferation and differentiation of mesoderm-derived cells. Cell Differ Dev. 1986;91:1–17.
Baird A, Esch F, Mormede P, Ueno N, Ling P, Bohlen P, Ying SY, Weherenberg W, Guillemin R. Molecular characterization of fibroblast growth factor: distribution and biological activities in various tissues. Recent Prog Horm Res. 1986;42:142–205.
Casscells W, Speir E, Sasse J, Klagsbrun M, Allen P, Lee M, Calvo B, Chiba M, Haggroth L, Folkman J. Isolation, characterization, and localization of heparin-binding growth factors in the heart. J Clin Invest. 1990;85:433–441.
Spirito P, Fu Y-M, Yu Z-X, Epstein SE, Casscells W. Immunohistochemical localization of basic and acidic fibroblast growth factors in the developing rat heart. Circulation. 1991;84:322–332.
Mima T, Ueno H, Fischman DA, Williams LT, Mikawa T. Fibroblast growth factor receptor is required for in vivo cardiac myocyte proliferation at early embryonic stages of heart development. Proc Natl Acad Sci U S A. 1995;92:467–471.
Engelmann GL, Dionne CA, Jaye MC. Acidic fibroblast growth factor and heart development: role in myocyte proliferation and capillary angiogenesis. Circ Res. 1993;72:7–19.
Schaper W, Ito W. Molecular mechanisms of coronary collateral vessel growth Circ Res. 1996;79:911–919.
Bernotat-Danielowski S, Sharma HS, Schott RJ, Schaper W. Generation and localisation of monoclonal antibodies against fibroblast growth factors in ischaemic collateralised porcine myocardium Cardiovasc Res. 1990;27:1220–1228.
López JJ, Edelman ER, Stamler A, Hibberd MG, Prasad P, Thomas KA, DiSalvo J, Caputo RP, Carrozza JP, Douglas PS, et al. Angiogenic potential of perivascularly delivered aFGF in a porcine model of chronic myocardial ischemia. Am J Physiol. 1998;274:H930–H936.
Harada K, Grossman W, Friedman M, Edelman ER, Prasad PV, Keighley CS, Manning WJ, Sellke FW, Simons M. Basic fibroblast growth factor improves myocardial function in chronically ischemic porcine hearts. J Clin Invest. 1994;94:623–630.
Lazarous DF, Scheinowitz M, Shou M, Hodge E, Rajanayagam S, Hunsberger S, Robinson WG Jr, Stiber JA, Correa R, Epstein SE. Effects of chronic systemic administration of basic fibroblast growth factor on collateral development in the canine heart. Circulation. 1995;91:145–153.
Unger EF, Banai S, Shou M, Lazarous DF, Jaklitsch MT, Scheinowitz M, Correa R, Klingbeil C, Epstein SE. Basic fibroblast growth factor enhances myocardial collateral flow in a canine model. Am J Physiol. 1994;266:H1588–H1595.
Banai S, Jaklitsch MT, Casscells W, Shou M, Shrivastav S, Correa R, Epstein SE, Unger EF. Effects of acidic fibroblast growth factor on normal and ischemic myocardium Circ Res. 1991;69:76–85.
Franz WM, Breves D, Klingel K, Brem G, Hofschneider PH, Kandolf R. Heart-specific targeting of firefly luciferase by the myosin light chain-2 promoter and developmental regulation in transgenic mice Circ Res. 1993;73:629–638.
Kostin S, Hein S, Bauer EP, Schaper J. Spatio-temporal development and distribution of the intercellular junctions in adult rat cardiomyocytes in culture. Circ Res. 1999;85:154–167.
Price RJ, Owens GK, Skalak TK. Immunohistochemical identification of arteriolar development using markers of smooth muscle differentiation: evidence that capillary arterialization proceeds from terminal arterioles. Circ Res. 1994;75:520–527.
Pu L-Q, Sniderman AD, Brassard R, Lachapelle KJ, Graham AM, Lisbona R, Symes JF. Enhanced revascularization of the ischemic limb by angiogenic therapy. Circulation. 1993;88:208–215.
Tabata H, Silver M, Isner JM. Arterial gene transfer of acidic fibroblast growth factor for therapeutic angiogenesis in vivo: critical role of secretion signal in use of naked DNA Cardiovasc Res. 1997;35:470–479.
Dilley RJ, Schwartz SM. Vascular remodeling in growth hormone transgenic mouse. Circ Res. 1989;65:1233–1240.